Business establishments have the potential to shape individual behaviors including dietary intake, physical activity patterns, and management of chronic medical conditions, all with relevance to risk of myocardial infarction and related cardiovascular disease (CVD) outcomes. Policy strategies targeting local businesses may play an important role in CVD prevention, extending years of healthy life, and facilitating independent living. We propose to bring together commercially available big data and detailed population-based cohort data to examine the role of local business environments in CVD and racial disparities in CVD as they emerged across decades. Longitudinal studies of local environments have been strongly recommended as a direction for future research to advance our understanding of causal neighborhood effects on health. The local density of businesses with potential relevance to CVD has changed over time, and such temporal fluctuations vary across regions, cities and neighborhoods. These differences can be used to investigate how residentially stable older adults are affected by their local context. Meanwhile, older adults who relocate to a new home address may respond to their new context by changing health behaviors, offering another window into causation. However, careful attention is needed to acute health changes that precipitate relocation, and which local environment features predict aging in place despite new health or functional limitations. In the proposed project, we will link longitudinal data to investigate how neighborhood change and residential relocation shape cardiovascular health and health disparities. We will use a census of national businesses, the National Establishment Time Series (NETS). NETS data for all 52.4 million US business establishments includes annual point-level geocodes and business characteristics for the years 1990-2012, offering an unprecedented opportunity to characterize changing business environments nationally at multiple spatial scales. Geographic context data (from NETS as well as longitudinal population, safety, and transportation data) will be linked to individual data from two ongoing cohort studies: the Cardiovascular Health Study (CHS) and the Reasons for Geographic and Racial Differences in Stroke (REGARDS) Study. CHS is a well-characterized population- based study of 5,888 older adults (100% age 65+ at baseline, 58% female, 16% African American). A majority of the CHS participants enrolled 1990-1993 from four US sites, with oversampling of African Americans, have been followed prospectively for the remainder of their lives, allowing in-depth characterization of neighborhood environments experienced by older adults in their final decades. REGARDS is also a prospective population-based cohort study with oversampling of African Americans, allowing for examination of racial disparities. However, the 30,239 REGARDS participants (49% age 65+ at baseline, 55% female, 42% African American) were enrolled more recently (2003-2007) and from a broader geographic region (48 contiguous US states, oversampling Stroke Belt residents). For these two cohorts with complementary strengths, cardiovascular events have been ascertained and longitudinal address data collected. This project brings together unparalleled data resources, builds on a track record of policy-relevant work with large geographic data resources, and leverages comprehensive cardiovascular cohort data.